Tumor growth and viability can be monitored by measuring the rate of glucose metabolism in the tumor, using fluorodeoxyglucose (FDG) positron emission tomography (PET). An existing NCI protocol studies subjects with liver tumors before and after therapy with tumor necrosis factor. The goal of this imaging physics project is to provide quantitative values of FDG uptake and metabolism by (l) developing and implementing an adequate kinetic model of glucose metabolism in normal and diseased liver, (2) investigating the errors associated with improper attenuation correction in studies of the abdomen, especially near the lung / liver tissue interface, and (3) permitting overlay of morphological MRI or CT images with the PET data. The first goal has been nearly achieved, although more testing is still necessary, and additional work is needed to be able to apply the models on a pixel by pixel (or at least on a small regional) basis. The second goal is complete. Work on the third goal has only recently begun. The initial attempts will utilize lower lung boundaries to achieve image alignment prior to overlapping the images.